Paul Wexler Tel-Aviv University Why There May Have Been Contacts between Slovenes and Jews before 1000 A.D. V okviru nove teorije o preseljevanju ~ i d o v v Evropo avtor trdi, da je najbolje gledati na jidiS kot na naretje 1uiiSke srbSCine, ki je bilo naposled preleksikalizirano s srednjevisokonemSkim besed- jem. Do tega preseljevanja na dvojeziCna IuiiSkosrbsko-nemSko govoreCa ozemlja, ki se je dogajalo do poznega prvega tisotletja, so se ~ i d j e selili prek Balkana, kjer so prevzeli enega ali vet balkan- skih jezikov: romanskega, grSkega iniali juinoslovanskega. Prispevek predstavlja dokaze v prid tej teoriji, s poudarkom na tistih, ki izhajajo s slovenskega ozemlja. In the framework of an innovative theory on the migration of the Jews to Europe, the origin of Yiddish is viewed as the result of relexification of a dialect of Sorbian to Middle High German vocabulary. Prior to their arrival in Sorbo-German speech territory by the late first millennium A.D., Jews migrated through the Balkans and shifted to one or more Balkan languages: Romance, Greek andlor South Slavic. The present paper presents evidence in support of this theory with an emphasis on evidence from the Slovene territory. Slovenia has never had a significant Jewish population, nor archaeological re- mains that attest to a Jewish presence in the pre-Slavic period, such as in Croatia, Serbia, Macedonia or Bulgaria; see, e.g., the synagogue at stip (Stobi), Macedonia from the 3rd-4th centuries or the mixed Avar-Jewish necropolis at Celarevo in the Vojvodina, thought to date from the 6th to 8th centuries.' Detoni claimed that Jews had settled in the vicinity of Maribor during the Roman period, but she gives no archaeological or historical evidence (1957-8: 72). In fact, her assertion that Jews appeared again in the Slovene lands in the 9th-10th century is also conjecture (ibid.). The first documented evidence of a Jewish presence in Slovenia dates from the 13th century, when Yiddish- and Italian-speaking Jews migrated south from Austria to Maribor and Celje, and east from Italy into Ljubljana. This is a good three cen- turies after the first mention of Jews in the Austrian lands. Valuable data on the Jewish presence in Slovenia comes from Johann Weichard Valvasor's compendious history of Carniola that appeared in 1689. Here we are told about the construction of a new synagogue in 1213 in Ljubljana (3: 1689: 710 - which implies the presence of a Jewish community perhaps in the 1100s), a document signed by Jews from Ma- ribor, Radkersburg, Judenburg, St. Veit an der Glan and Ljubljana at the close of the 13th century (ibid., 4: 309-lo), the claims that Jews poisoned wells in Ljubljana in 1337 (ibid., 4: 319), and that in 1496 they ridiculed the Christian sacraments and killed many Christian children for their blood, which was needed for the production of matzah (unleavened bread eaten at Passover) (ibid., 4: 393). ' On ~ e l a r e v o , and other cites, see Nedomatki 1988a; for a detailed discussion of Celarevo, see BunardiiC 1980 and Wexler 1993: 30, 193-95. Kollautz and Miyakawa 1970: I1 note that both Jews and Avars deposited objects in the grave of the deceased. P. Wexler. Whv There Mav Have Been Contacts between Slovenes and Jews before 1000 A.D. 57 Perhaps the most illustrious Jew to have resided in Slovene territory was the Ash- kenazic (Yiddish-speaking) rabbi Israel ben Ptaxya Isserlein (Iserlin), who was born in Regensburg in 1390; he was educated in Wiener-Neustadt, arriving in Maribor in 1427, where he apparently resided until 1445. He was the author of an important Jewish legal treatise in Hebrew, known as Trumat hadeien (Venice 1519), and died in 1460.' Jewish settlement in Slovenia was interrupted when the Jews were expelled from Styria and Carniola in 1496 and from Ljubljana in 15~15.~ Many of the expelled Jews migrated to Triest, Vienna and Sopron." The Italian Jewish family name Mor- purgo may come from Marburg, the German name for Maribor. Alternatively, the Jewish family name could be derived from the German city of Marburg in Hesse, which also had a Jewish settlement in the 14th century. The fact is that a number of Ashkenazic Jews in Italy named themselves after the German towns and areas from which they originally hailed; see, e.g. Luzzatto < German Lausitz (Ober- and Niederlausitz are the two districts in southeast Germany in which a small Sorbian- speaking population still resides). While large groups of Iberian Jews entered the South Slavic lands and the Ottoman Empire after 1492 (the year of their expulsion from the Kingdom of Spain), the bulk of these settlements were established in Al- bania, Bosnia, Bulgaria, Macedonia, Greece, and Turkey, and later on in Serbia and Croatia; Slovenia never attracted a significant Judeo-Spanish-speaking population. Despite the lack of material evidence of an early Jewish presence in Slovene lands, I believe that it is possible to propose, cautiously, the innovative thesis that there were Jews in this area of the Balkans during the first millennium. Contacts between Jews and Slovenes at this time must be established almost exclusively by linguistic evidence (primarily by the facts of Yiddish, and secondarily by the Slavic languages in contact with the latter and possibly other Jewish languages). I theorize that the Balkan Jews of the first millennium consisted of a minority ethnic Palesti- nian component and a majority population from Europe and Asia Minor which had become either receptive to Jewish culture or who had actively converted to Judaism. In my view, it is among Balkan Jews who migrated north into the mixed Ger- mano-Sorbian lands (of present-day eastern Germany) that the Ashkenazic (North European) Jewish community developed by the latest in the 9th-10th centuries. See Wenninger 1995a: 836. For a list of names of Jews recorded in Maribor from 1350 on, see ibid., 835-7. For a description (with photographic illustrations) of seals belonging to Slovene Jews in the 14th-15th centuries, see Friedenberg 1987: 297-302. Some of the names of Jews as re- corded on their seals and in contemporaneous Christian documents differ from Ashkenazic names known from monolingual German lands, see, e.g. Maul (Maribor 1482, possibly a contraction of Hebrew Srnu'el 'Samuel'?). On Slovene Jewish names, see also the discussion below. Iserlin's Yiddish was the subject of study by Lunski 1924 and Weinreich 1924. Iserlin's writ- ings include the term pjnkwWf - pjnkwWs 'Christian Pentecost', which may be of Hungarian or Slovene origin (though ultimately from German, see my 1987: 33-4), e.g., Hungarian piinkiisd, Slovene binkofti. Iserlin's spelling with " indicates that the term was regarded as a foreignism - i.e. non-Germanism and non-Hebraism. For details of Jewish settlement in Celje, Judenburg, Maribor, Radkersburg and St. Veit an der Glan, see Tykocinski 1934, Detoni 1957-8, Lazarus 1968, NedomaEki 198813, NedomaEki and Goldstein 1988 and Wenninger 1987a, b, c, 1995a, b, c. Further bibliography on Ashkenazic Jewish history in Slovene territory is cited by Detoni 1957-8 and Wenninger 1987c: 698-702 and 1995a. See Detoni 1957-8: 74. 58 Slovenski jezik - Slovene Linguistic Studies 1 (1997) The evidence of contact between the Slovenes or a non-Slavic population inhab- iting the historic Slovene lands and the Jews in the late first millennium is scant. The failure of historians to pose the possibility of such historic contacts has resulted in insufficient exploitation of the few known facts, as well as a reluctance to seek out new evidence. Hopefully, more research will permit greater precision of formula- tion and corroboration. Moreover, when set within the broader framework of the Balkans as a whole - where somewhat more historical, archaeological and linguistic evidence is available - an early Slovene-Jewish connection is not implausible. De- tailed evidence of a Balkan substratum (Balkan Latin, Greek, and Slavic) in Yiddish was presented in Wexler 1992. The traditional view of Yiddish genesis is that French and Italian Jews migrat- ed to the Rhineland and secondarily to Bavaria in the 9th century; instead of adopt- ing the local German dialects, the Jews adapted German linguistic material, laying the basis for a unique Judaized German, later known as Yiddish. Incorporated into Yiddish was a small, unique Romance component, unknown in the coterritorial German dialects; most of the Romance component was derived from Old French (specifically to a Judaized form of the latter), with a smaller corpus from Judeo-I- talian.' In this model, Yiddish is defined as a Judaized form of High German, and the Slavic component, which is extremely marked in all aspects of Yiddish, is re- garded as a relatively late acquisition. In contemporary Eastern Yiddish dialects, the Slavic component reaches approximately 10% of the lexicon; a further 15% of the lex- icon is of Hebrew-Aramaic origin - the two liturgical languages of the Jews - and the remainder German. For reasons that cannot be repeated here for lack of space, I suggested (in 1991 and 1993) that a Germanic genetic assignment for Yiddish was unconvincing, since an analysis of the syntax and phonology of Yiddish strongly suggested that the latter was originally a Judaized form of Sorbian which became relexified to High German be- tween the 9th and 12th centuries (on Judeo-Slavic in general, see my 1987). Despite the relexification to mainly High German vocabulary, the syntax, phonotactics and pho- nology of Yiddish have remained to this day West Slavic. However, due to the exist- ence of its overwhelming High German lexicon, Yiddish has given the impression, to native speakers and non-observers alike, of being a form of High German. My conception of Yiddish genesis differs from the traditional view in three important ways: (a) I believe that Jews of Balkan origin, speaking South Slavic, Greek, and/or Balkan Romance, migrated into the (south)east German lands by the 9th century, at a time when these lands were still predominantly Sorbian-speaking! The traditional view is best presented in Weinreich 1973 and Birnbaum 1979. Mieses deserves the credit for first suggesting that some Slavisms in Yiddish might be of Balkan origin (1924: 232-3, 271). This is a plausible suggestion, given the Balkan Romance sub- stratum in Yiddish (see details in my 1992), but probably impossible to prove, since many Yiddish Slavisms could, in principle, be derived from any number of Slavic languages. Note that the Ash- kenazic practice of baking braided bread for the Sabbath and other festive occasions appears to be of Greco-Roman origin; if so, it may have been acquired by Jews in the Balkans and brought north to the German lands. It survived in the new Ashkenazic homeland, since the custom was al- so practiced by Germans and West Slavs. Thus, the Yiddish name for the bread, xale < Biblical Hebrew hallah 'piece of unbaked dough given to the priests as a tithe', looks like an attempt to P. Wexler. Whv There Mav Have Been Contacts between Slovenes and Jews before 1000 A.D. 59 (b) In the Germano-Sorbian lands, the Balkan Jews settled among Slavic speakers and became speakers of Sorbian. As the German Drang nach Osten gained momentum in the 9th century and Germans became the majority in the formerly Slavic territories, Sorbian (and Polabian) became replaced by German. In the case of most of the pagan Sorbs and Polabians, the shift to German was complete; that is, Sorbs and Polabians became bilingual in German, and then eventually gave up their original Slavic speech. Concomitantly, they also became Christians. The Sorbian Jews, on the other hand, were more resistant to German language, culture, and re- ligion (Christianity), and thus made only a partial language shift, which involved a replacement of their lexicon, while the syntax and phonology remained largely in- tact. The tiny unique Romance and Greek component that remains in all dialects of Yiddish is most smoothly derived from north Italian, Friulan, and Balkan Romance dialects; a possible French contribution is evident only in a handful of Yiddish an- throponyms (see, e.g., Yiddish bunem 'male name' < French bon homme or Italian buon uomo literally 'good man').7 (c) Ashkenazic ethnographic, linguistic and historical evidence all suggests that the bulk of the Balkan Jews who eventually migrated north into the mixed Ger- mano-Sorb lands around the 9th century were not of ethnic Palestinian Jewish ex- traction. Rather, the European Jews consisted overwhelmingly of an indigenous pro- selyte community and a minority of Palestinian Jews. The demographic structure of the early European Jews accounts for the loss of spoken Hebrew and Aramaic in Europe very soon after Palestinian Jews arrived in Europe, and the absence of He- brew and Aramaic literature (including the original texts of the Bible and Talmud) in Europe (and North Africa) until the 10th century. The evidence for widespread conversion to Judaism in the Hellenic world, the Balkans and Sorbian lands was laid out in Wexler 1991 and 1993. There is good reason to believe that the Southern and Central Europeans who swelled the ranks of the Jewish community included a small Turkic population. The Slovene lands may have well been a part of the territory in which the conversions to Judaism took place, before the second millennium, when Christianity became insti- tutionalized. One dramatic piece of evidence for tbe claim of a Turkic component comes from the mixed Avar-Jewish necropolis in Celarevo (Vojvodina) that I men- tioned above. Here, a number of Jewish graves were identified by Jewish decorations and an occasional Hebrew inscription on the stones. The argument that some Avar invaders to Central and Southern Europe prac- ticed Judaism (which they acquired in the mixed-Turkic Khazar lands - where Ju- daism became the state religion in the 8th century?) is further strengthened by the South Slavic and Rumanian tradition of associating Jews with legendary giants as well as by the West and South Slavic and Greek tradition of associating Avars with giants. Judaize the pre-Christian German name of similar phonetic shape, which was derived from Holle, the name of the witch to whom the braided bread was traditionally dedicated (for details and rich bibliography, see my 1993: 115-8). Alternatively, the Ashkenazic Jewish braided bread might have been first acquired on German territory, since the practice of baking this sort of festive bread is well known among Germans (in rural areas especially). Where it exists today in West Slavic areas, the bread is called either by this or another Yiddish "Hebraism," e.g., barxes. On the unique Romance and Greek components in Yiddish, see Wexler 1992. 60 Slovenski iezik - Slovene Linguistic Studies 1 (1997) The relevant data are as follows: (a) 'giant' is formed from the term for 'Jew': North Bulgarian, Eastern Serbian iid, Bulgarian iidovec, diidavec; Rumanian jidov: (b) 'Jew' (substandard) is formed from the term for 'high, tall': Macedonian vi- sok ' J ~ W ' ; ~ (c) 'giant' is formed from the term for 'Avar': Slovene oberI0, Czech obr , Old Polish obrzym (Modern Polish olbrzym), Upper Sorbian hober , hobjer', Byzantine Greek dvrimos. Some scholars have proposed that the term "Jew" became the basis for the term "mythical giant" since features such as abnormal tallness and a mythical pre- historical setting could be readily ascribed to an "alien" or non-indigenous element of the population." Given the fact that Jews and Avars were in contact (to judge from the important finds at the Avar necropolis at Celarevo and the tradition of de- riving a Carniolan population from the Biblical "Ashkenazim" - see below), it may be that Jews and Avars were given the status of giants since there were adherents to Judaism among the Avar invaders of Central Europe and the Balkans in the 5th-6th centuries. Hence, the specific association of giants with Jews could have Saineiinu noted that Rumanian used both %tar 'Turk' and Jidov 'Jew' to denote giants, a fact which he ascribed to the conversion of the Khazar Turks to Judaism (1888: 154, 158, 160). In Sieben- biirgen (Rumania) giants or Jews were allegedly the builders of fortresses, while in Bulgaria, Jews and Tatars performed this function (Winkler 1930-1: 1649). The Bulgarian and Rumanian data are dis- cussed by the Biilgarski etirnologiten retnik 1 (1971): 543-4 and Matl 1956: 294, 306. For Macedonian, see Giibjuv 1900: 846. Jews and Avars are not alone in ascending to the status of giants; in Southern Bulgarian and Macedonian, giants are designated by the term elini literally 'Greeks' (see Matl 1956: 303, 306). A semantic development parallel to the Balkan example seems to characterize Ukrainian as well, if Ukrainian kozarljuha 'strong, courageous warrior' is derived from kozar 'Khazar' (see Pritsak 1963: 930); a pejorative development from the ethnic name Khazar is Hungarian kazdr 'kike'. For further discussion, see Wexler 1987: 142. Another Slavic example of an ethnic term assuming the meaning of giant is Croatian, Russian ispolin, Pol- ish stolin, stolyrn, stwolin etc. 'giant' from the name of an ethnic group known as Spalaei in Pli- ny (see Niederle 1924: 64, 82-3; Vasmer 1953-8; Machek 1956; Skok 1: 1971: 731). lo It is interesting that Slovene is the only South Slavic language in this listing. See Vasmer 1953-8; Kollautz and Miyakawa 1970: 168, 171; Skok 2: 1972: 535; Wexler 1993: 196-7. Machek 1956, 1971 rejects the connection between 'giant' and 'Avar'. The Alans, ancestors of the Iranian Ossetes, have provided the Karaites (a Turkic-speaking Judaic sect in the Crimean Peninsula, western Belorussia, western Ukraine, and Lithuania) with their word for giant, alankasar (see Moskovich and Tukan 1985: 90). Note also that Romani biboldo 'Jew' (from the root for 'unbap- tised') appears throughout German slang (Rotwelsch) in this meaning, but in Berlin speech, ba- bold(e) is cited from 1812 in the meaning of 'big, strong man' (see Schildt and Schmidt 1986: 349 and Wolf 1956: 450). It would be interesting to know if the Gypsies acquired this new meaning of biboldo in the Balkans. Many Western European Gypsies are descended from groups which resid- ed previously in the Balkans for several centuries. ''See Iliev 1890: 189-98, Mazon 1932, Biilgarski etirnologiten retnik 1: 1971 373, 543-4, 11: 1979: 89, Ivanov and Toporov 1980: 37, Pareba 1993, Lozanova 1994: 40 and Wexler 1993: 186-7. For the view that Jews were not intended by the Russian term Zidovin, the name of a giant in the Russian byliny, see Mollova 1973: 101-2 and Vasil'ev 1995: 18-9. On the tendency to make non-existent peoples into giants, see Kollautz and Miyakawa 1970: 168, 171. P. Wexler. Whv There Mav Have Been Contacts between Slovenes and Jews before 1000 A.D. 61 been based on an actual historical event. It is inconceivable that Jews, lacking an army or autonomous political structure (apart from the Khazar Empire in the Pon- toCaspian steppe area, destroyed by the Kingdom of Rus' in the 10th century), could have been transformed into giants in the local folklore, had there been no connection between Jews and the warlike Avars. Most of the territories in which the term for Avar assumed the secondary meaning of giant were occupied by the Avars." Further support for the link between the Jews and Avars is that in the Austrian Steiermark there are several towns based on the ethnonym "Avar" located near a town called Judenburg (the name is attested since about 1080), see, e.g., Oberweg (1348 ff.) < Common Slavic *obwovit'i - Old Church Slavic obwin.6 'Avar' (for this and other toponyms based on "Avar", see Mader 1986: 42, 67, 112). Even if Kronsteiner's alternat- ive derivation of Judenburg from a Slavic root independent of Jews is correct (see his 1978: 153), the question still remains of why the placename eventually became reinter- preted as "town of the Jews"; due to a local tradition that associated Avars and Jews? Curiously, one further term which combines the semantic primitives of "Jew" and 'large' is "Hebrew" gol~m 'supernatural male servant made out of clay whose purpose is to protect the Jews'. The term is traditionally derived from Hebrew 'amorphous substance' but I suspect that the Hebrew connection is the result of sub- sequent folk etymologizing. The real etymon is Slavic - either Old Czech holomek (zemsky') 'servant, judge's or executioner's helper', Old Polish holomek 'house servant' (first attested in the mid-15th century), or Czech holemy' 'strong' (< Common Slavic *goltm% large'; see also dialectal Slovene golmei 'very big'). The earliest Jewish ac- count of the story of a golem-servant seems to be in the 13th-century Hebrew writ- ings of German Jewish pietists and mystics, but since the story of a supernatural en- tity has non-Jewish Byzantine roots, we may want to posit a Balkan, Slavic or Greek origin for the Jewish story and "Hebrew" gol~m.'~ Particularly important and, to the best of my knowledge, ignored in the litera- ture is Valvasor's mention of a local tradition that three Carniolan tribes, the Cimb- ri, Galatae and Teutones, were descended from the Ashkenazim (ibid., 1: 1689: 65, 2: 41-2, 45-9). This ethnonym is derived from the biblical Hebrew toponym aika- naz, apparently located in Iranian speech territory, by which medieval European Jews first designated Slavs, then Germans, and finally Germans and German Jews. Valvasor rejects this claim as unfounded, as well as the claim that Hebrew aikanaz might be related to the Saxon dynastic name A~kanien?~ The existence of such a tradition suggests that the Avar conquerors of the 5th-7th centuries may have in- cluded a significant number of Jews in their midst. The tradition that Valvasor referred to finds some further support in the asso- ciation of the Balkan Romance populations and the Ashkenazic Jews. Romance- speaking populations in the Balkans (Arumanians, Meglenorumanians, Rumanians) l2 See Kronsteiner 1978: 143, fn 28. Old Russian participation in this isogloss probably has to be attributed to West Slavic influence. l3 For details, see Skok 1934: 76, Zareba 1965, TrubaEev 1979, Rusek 1983 and my 1993: 132-3. l4 Ibid., 2: 48-9. For a discussion of the Hebrew term afkanaz and the possible connection with Askanien, see my 1987: 160 and 1991: 48. Until recently I was unaware of Valvasor's discus- sion of these topics. 62 Slovenski iezik - Slovene Linguistic Studies 1 (1997) are historically known by the epithet Vlax in Slavic languages. This ethnonym orig- inally denoted a Celtic tribe, known in Latin as the Volcae (an ethnonym related to English Welsh, [Cornwall]. In older forms of German, Walsche denoted both Celtic and Romance speakers (Andree 1874: 134). The term has become the basis of Ger- man Rotwelsch 'slang', literally 'beggar's incomprehensible (Romance!) language'. The Volcae were a partly Romanized Celtic tribe from the Middle Danube, some of whom migrated to Languedoc (between the Rhone and Toulouse). At least since the late Middle Ages, Germans, Slavs and Greeks have applied the term vlax to Roman- ized populations in their midst and sometimes even to other populations who adop- ted a pastoral or nomadic way of life, which was characteristic of the Romance minority. It is curious that the same root has become a common Ashkenazic Jewish (Yiddish) family name, in two forms, valax and blox (usually spelled Wallach and Bloch in Latin-alphabet languages, following German orthographic norms). The as- sociation between a Celtic tribal name in the Balkans (with Romance connections) and the Jews may be quite old. Note also that archaic Bulgarian dirmon 'leather grain sieve with large holes' surfaces in dialects with a variety of ethnic associations - Jews, Vlax and Roma (Gypsies), see, e.g., dirmon barba 'Jew', dirmonka 'Vlax woman', dirmule 'small Vlax' (the dialects where these forms are known are un- specified by the Bilgarski etimologiten retnik I : 1971, which notes only that the Vlax were known to be specialists with sieves). It is unclear if this association of Jews and Vlax is in any way the basis for the Yiddish names cited above. One final piece of evidence of Turko-Jewish contacts in the first millennium comes from the naming patterns of the Ashkenazic Jews themselves. While the practice of naming a child after the holiday on which it was born is found in many different cultures (see French Noel, Noelle, Italian Pasquale - though nowadays the use of these names is no longer motivated by the fact of being born on the respec- tive holidays), it seems to be particularly popular in Turkic speech communities. It is interesting to note that the use of the Hebrew terms p~sah 'Passover' (the basis of Church Slavic pasxa) and hanukkiih 'Hannukah' as boys' names are found historical- ly among Jews only in the North European area and in the 10th-century Ukrainian Khazar community. Up until the 13th century, the European territory covered by these two names included the West and East Slavic territories and the eastern Ger- man lands, approximately up to the limes sorabicus which was the dividing line be- tween the monolingual German lands of the Holy Roman Empire and the bilingual Germano-Slavic areas in the 9th century (see also Wexler 1987). Documents from the late 14th century from the Slovene area preserve examples of both of these Hebrew names, see, e.g., Paskul a Jew from Triest resident in Ma- ribor and Hankein (= Xanuka) also from Maribor (see Wenninger 1995a: 836-37). Either this naming tradition reached the Slovene lands with the arrival of Ashkenaz- ic (Yiddish-speaking) Jews in the 13th century, or diffused to this area even earlier with the Judeo-Turkic migrations into the Balkans. In addition to putative Turkic components in the Ashkenazic Jewish ethnogene- sis, there is also a possibility of Iranian elements as well. Consider Yiddish iabaS 'tip given to the musicians at a wedding by guests who participate in the dancing' < Iranian SiibiiS with the same meaning. The Iranianism surfaces in the slang regis- ters of a number of West and East Slavic languages, Hungarian, Dutch and German in both a full and truncated form in the meaning of money (see details in Wexler P. Wexler. Whv There Mav Have Been Contacts between Slovenes and Jews before 1000 A.D. 63 1987: 64-69, 1993: 110). The practice of tipping musicians at a wedding is attested among the Polabian Slavs in western Germany (Polabian became extinct in the Lu- neburg region at the beginning of the 19th century), Macedonians (e.g., at PeSteni: see Leibman 1972: 128), as well as among western Hungarians (though in the latter community the tip is given to the church primate; see Rottleuthner 1985: 40). I do not know if either the term or practice was ever attested on Slovak, Croatian or Slo- vene territory!' One final connection between Jews and Iranians involves the Balkan term for 'werewolf', e.g., Macedonian Greek vroukdlakas, Cakonian Greek vourkdlakas, Rumanian valcolac, Albanian vurvollak, Serbo-Croatian vukddlak, Slovene volko- dlak. The term (which Nichols believes comes ultimately from Iranian, and not from Slavic, as has been traditionally claimed) surfaces in early 13th-century German He- brew texts (it is totally unknown in coterritorial German dialects, or in Yiddish it- self, for that matter). This fact suggests that the German Jews were heirs to Balkan terms and traditions unknown among the coterritorial Germans; the best explanation is that these terms and traditions were brought to Germany by Balkan Jews (for de- tails, see Vlachos 1971: 226, fn 48 and my 1993: 105-6). This may not be the only example of a linguistic or ethnographic Balkanism among German Jews. Another possible linguistic example might be the Yiddish expression fojglmilx literally 'bird's milk,' which has the meaning of something unusual, very precious, i.e., something unattainable or nonexistent. This expression is unattested in German dialects, but is widespread in the Slavic languages, as well as in Greek, Turkish, Arabic, and Mod- ern Hebrew (where it is probably a loan from Yiddish); see Slovene ptitje mleko (now in this meaning or only as a botanical term, 'Ornithogalum'?), Bulgarian ptite mljako. In South Slavic languages, or in East Slavic languages, due to the impact of Old Church Slavic, we can imagine a Greek source; Czech ptati mltko or Polish ptasie mleko on the other hand, are more problematic, given the absence of a sig- nificant Church Slavic imprint on West Slavic languages. It is not inconceivable that the Yiddish expression owes its origin to a Balkan Jewish language, and is the direct source of the expression in West Slavic languages. Finally, there is evidence from the 15th century that Slovene Jewry either com- prised a non-Yiddish-speaking minority, or developed on a non-Ashkenazic Jewish substratum. For example, in a German document from Kranj, dated 1446 and signed by five Jews, the 11th month of the Hebrew calendar, Sva, is transliterated as Sam- bat (Valvasor 4: 1689: 310, cited also by Friedenberg 1987: 297 without comment). Yiddish pronunciations of Hebrew retain Hebrew S as such (except in the Belarusian and Lithuanian dialects of Yiddish which have sibilant confusion, and hence are ir- relevant to the present discussion), e.g., Hebrew SabbiLt > Yiddish Sabes 'Sabbath, Saturday'; nor is there any justification for Hebrew v becoming mb. It appears that the name of the Hebrew month was confused (by Christian scribes?) with Balkan Slavic or Judeo-Greek reflexes of Hebrew SabbiLt 'Saturday, sabbath'. In Judeo-Greek, geminated b was dissimilated to mb due to the absence of the former in the sound pattern; I assume only Jews would have been sensitive to the need to capture somehow the geminated consonant in the original Hebrew. 150n Iranian populations in the East Balkans in the early Christian era, see Du Nay 1977: 189-90, 201-2. 64 Slovenski jezik - Slovene Linguistic Studies 1 (1997) Compare Judeo-Greek sambata vs. standard Greek sabbata, sabbaton 'Saturday, Sabbath', with distinctly "non-Hebraic" forms. The Judeo-Greek reflex has a limited diffusion in neighboring Macedonian dialects, see Southeast Macedonian sambuta (northeast of Saloniki) Southwest (Kostur, Kor~a) Macedonian samba vs. standard Macedonian sabota (see Vidoeski 1983: 130, 132 and further discussion in Wexler 1987: 24-27). There is no m in the Sl~vene,'~ Serbo-Croatian or Bulgarian cognates, however, though neighboring non-Slavic languages do follow the Judeo-Greek pro- nunciation norm, see, e.g., Hungarian szombat, Rumanian s6mbLtL. The replace- ment of S by s could have also developed in a (Judeo-)Greek-speaking milieu. Note also that in the western-most Iranian language, Ossete, Hebrew SabbiLt surfaces as sabat 'Saturday, sabbath'. Ossete has a few Hebrew, Aramaic, and Judeo-Greek ele- ment~!~ The example of Sambat from the Slovene texts suggests that a South Balkan pronunciation norm of Hebrew may have been diffused to the northern-most Bal- kans. This is valuable evidence in support of the thesis that the Ashkenazic Jews are descended in the main from migrants from the Balkans. I believe that the evidence both presented here and referred to in other studies justifies a fresh look at Jewish migrational patterns to and within early medieval Europe. My findings suggest that the Ashkenazic Jews are in large part descended from a Slavo-Turkic proselyte population rather than from transplanted Palestinian Jews. Slavo-Turks may have become Judaized in the ethnically mixed Khazar lands (in the Ponto-Caspian steppes) andlor in the Slavo-Avar settlements in the Balkans. If the former, then the first Jews in the Balkans were very likely of Asian Turkic rather than Mediterranean Jewish origin, and were part of the Drang nach Westen of Slavs, Avars, Magyars, and others, which was gathering momentum beginning with the 6th century. If the latter, then the first Jews in the Balkans were to some extent of indigenous extraction. In any case, the German Jews could not be the de- scendants of French and Italian Jews, as is commonly believed. Once ensconced in the German lands in the early second millennium, some of the Jews subsequently migrated south into northern France where they assisted in the formation of major Jewish cultural centers, though most German Jews joined the Germans in the latter's Drang nach Osten into monolingual Slavic territories, where they were to coexist un- til World War 11. The Slovene lands were probably part and parcel of the broader Balkan context in which Ashkenazic Jewry - the largest branch of the so-called di- aspora Jews - was ~reated!~ l6 [The Slovene evidence is as yet ambiguous. Common Slavic *o and *Q have merged in Slo- vene, rendering the recovery of the form at the time of borrowing problematic. Upper Carniolan zbdta (< *subota < *sobota (?)) as well as Inner Carniolan sabaota point to the possibility of *o; on the other hand, the place name Murska Sobota is attested as Murazornbata (1348), Murazurnba- ta (1365), and Mura Schornbath (1423), indicating a nasal element (see Bezlaj 1995: 282). - Op. ed.] l7 See Wexler 1987: 21, 222, 226-27. On Ossete, see also footnote 10 above. l8 Future research should explore why Slovene lacks Jewish expressions that surface in neigh- boring languages among Christians, e.g., the Judeo-Greek practice of denoting the holiday of Yom Kippur (Day of Atonement) as the "great day" surfaces in Hungarian as hosszanap and in Czech as dlouhy' den; there is apparently no trace of this term in Slovene. Slovene also has no reflex of Judeo-Greek rn in the word for Saturday. P. Wexler. Whv There Mav Have Been Contacts between Slovenes and Jews before 1000 A.D. 65 References Andree, R. (1874). Wendische Wanderstudie. Zur Kunde der Lausitz und der Sor- benwenden. Stuttgart. Bilgarski etimologiten retnik (1971-95). Vols. I- IV. Sofija. Bezlaj (1995). F. Bezlaj Etimoloiki slovar slovenskega jezika III. Ljubljana. Birnbaum, S. A. (1979). Yiddish. A survey and a grammar. Toronto - Buffalo. BunardiiC, R. (1980). Izloiba menore iz celareva. Beograd. Detoni, M. (1957-8). Mariborski geto. Jevrejski almanah 72-4. Du Nay, A. (1977). The early history of the Rumanian language. Lake Bluff, Illinois. Friedenberg, D. (1987). Medieval Jewish seals from Europe. Detroit. GZbjuv, P. K. (1900). Prinos kgm bglgarskite tajni ezici. Sbornik za narodni umotvo- renija, nauka i kniinina 16-17: 842-75. Iliev, At. (1890). Bglgarskite predanija za ispolini, nareteni elini, iidove i latini. Sbornik za narodni umotvorenija i narodopis 3: 179-205. Ivanov, V. V. and V. N. Toporov (1980). 0 drevnix slavjanskix knonimax. (Osnov- nye problemy i perspektivy.) Slavjanskie drevnosti. ~ t n o ~ e n e z , material'naja kul'tura drevnej Rusi, 11-45. Kiev. Kollautz, A. and H. Miyakawa (1970). Geschichte und Kultur eines volkerwande- rungszeitlichen Nomadenvolkes. Die Jou-Jan der Mongolei und die Awaren in Mitteleuropa 2. Die Kultur. Klagenfurt-Bonn. Kronsteiner, 0 . (1978). Gab es unter den Alpenslawen eine kroatische ethnische Gruppe? Wiener slavistisches Jahrbuch 24: 137-57. Lazarus, F. (1968). Marburg. Germania judaica, ed. Z. Avneri, 212: 522-3. Tiibingen. Leibman, R. (1972). Wedding customs in the Ohrid village of PeSteni. Makedonski folklor 9-10: 125-40. Lozanova, G. (1994). The category of "alien-one's own" in folk culture. ~ t u d e s bal- kaniques 30: 37-42. Lunski, X. (1924). Iserlins jidig. Jidiie filologje 288-97. Warsaw. Machek, V. (1956). Expressive Vokaldehnung in einigen slavischen Nomina. Zeit- schrift fur Slawistik 1 (4): 33-40. - - (1971). Etymologicky' slovnik jazyka tesktho. Prague. Mader, R. (1986). Die Alpenslawen in der Steiermark. Eine toponomastisch-archao- logische Untersuchung. Vienna. Matl, J. (1956). Zur Bezeichnung und Wertung fremder Vijlker bei den Slaven. Fest- schrift fur Max Vasmer zum 70. Geburtstag, eds. M. Woltner and H. Brauer, 293-306. Wiesbaden. Mazon, A. (1932). Svjatogor ou Saint-Mont le gkant. Revue des ttudes slaves 12: 160-201. Mieses, M. (1924). Die jiddische Sprache. Berlin - Vienna. Mollova, M. (1973). Quelques lexemes turks septentrionaux en - t - j ... dans les langues slaves mkridionales. Zeitschrift fur Balkanologie 9 (1-2): 89-127. 66 Slovenski jezik - Slovene Linguistic Studies 1 (1997) Moskovich, W. and B. Tukan (1985). Caraimica. The problems of the origin and his- tory of East European Khazars in the light of linguistic evidence. Slavica hie- rosolymitana 7: 87-106. NedomaCki, V. (1988a). ArheoloSki nalazi. ~ i d o v i na tlu Jugoslavije, 17-25. Zagreb. - - (1988b). Geta i mahale. Ibid. 27-31. - - and S. Goldstein (1988). Jevrejske opCine u jugoslavenskim zemljama. Ibid. 113-41. Nichols, J. (1987). Russian vurdalak 'werewolf' and its cognates. Language, literature, linguistics. In honor of Francis J . Whitfield on his seventieth birthday, March 25, 1986, eds. M. S. Flier and S. Karlinsky, 165-77. Berkeley. Niederle, L. (1924). Slovanskt staroiitnosti. PJvod a potdtky SlovanJ vjchodni'ch 1 (4). Prague. Pareba, M. (1993). Otraienie staryx 5tniEeskix i jazykovyx kontaktov na Balkanax v bolgarskix i serboxorvatskix dialektnyx nazvanijax nekregtennogo rebenka. Balkanskoe ezikoznanie 36: 275-80. Pritsak, 0 . (1963). Altaic elements. Ukraine. A concise encyclopaedia, ed. V. Kubijo- vyC, 1: 928-31. Toronto. Rottleuthner, W. (1985). Bauerliches Hochzeitsbrauchtum in Sudmahren zur Jahr- hundertwende (Prittlach 1895). ~sterreichische Zeitschrift fur Volkskunde n. S. 39: 25-44. Rusek, J. (1983). Bulgarskie goltm.6 'wielki, magnus'. Rocznik slawistyczny 43: 21-28. SaineZnu L. (1888). Jidovi sZu iutari sZu uraisii. Anuar pentru Israeligi 11: 153-60. Schildt, J. and H. Schmidt (eds.) (1986). Berlinisch. Geschichtliche Einfuhrung in die Sprache einer Stadt. Berlin. Skok, P. (1934). Dolazak Slovena na Mediteran. Split. - - (1971-4). Etimologijski rjetnik hrvatskoga ili srpskoga jezika 1-4. Zagreb. TrubaEev, 0 . N. (ed.) (1979). ~t imolo~i tesk i j slovar' slavjanskix jazykov. Praslavjan- skij leksiteskij fond 6. Moskva. Tykocinski, H. (1934). Laibach. Germania judaica, eds. I. Elbogen, A. Freimann, H. Tykocinski, 1: 152. Breslau; reprinted Tubingen 1963. Valvasor, J. W. (1689). Die Ehre des Hertzogthums Crain und historisch-topographi- sche Beschreibung 1-4. Translated into German by E. Francisci. Ljubljana. Vasil'ev, M. A. (1995). aXors iidovin~ i drevnerusskoe jazyteskoe boiestvo v konteks- te problem Khazaro-slavica. Slavjanovedenie no. 2: 12: 21. Vasmer, M. (1953-8). Russisches etymologisches Worterbuch 1-3. Heidelberg. Vidoeski, B. (1983). A survey of the Macedonian dialects. In B. Koneski, A histori- cal phonology of the Macedonian language, 117-34. Heidelberg. Vlachos, T. (1971). Geister- und Damonenvorstellungen im siidosteuropaischen Raum griechischer Sprachzugehorigkeit. ~sterreichische Zeitschrift fur Volkskunde. n. S. 25: 217-48. P. Wexler. Whv There Mav Have Been Contacts between Slovenes and Jews before 1000 A.D. 67 Weinreich, M. (1924). Onmerkungen [Notes to Lunski 19241. Jidiie filologje, 297-302. Warsaw. - - (1973). Geiixte fun der jidiier sprax 1-4. New York. Wenninger, M. (1987a). Cilli. Germania judaica, ed. A. Maimon, 311: 206-10. Tu- bingen. - - (1987b). Judenburg. Ibid.: 592-6. - - (1987~). Laibach. Ibid.: 698-702. - - (1995a). Marburg. Germania judaica, eds. A. Maimon, M. Breuer and Y. Guggenheim, 3(2): 832-46. Tubingen. - - (1995b). Radkersburg. Ibid.: 1163-8. - - (1995~). St. Veit an der Glan. Ibid.: 1303-4 Wexler, P. (1987). Explorations in Judeo-Slavic linguistics. Leiden. - - (1991). Yiddish - the fifteenth Slavic language. International Journal of the sociology of language 91: 1-150, 215-25. - - (1992). The Balkan substratum of Yiddish. (A reassessment of the unique Ro- mance and Greek components). Wiesbaden. - - (1993). The Ashkenazic Jews. A Slavo-Turkic people in search of a Jewish identity. Columbus, Ohio. Winkler (1930-1). Heiden. Handworterbuch des deutschen Aberglaubens, eds. E. Hoffmann-Krayer and H. Bachtold-Staubli, 3, columns 1634-53. Berlin - Leipzig . Wolf, S. A. (1956). Worterbuch des Rotwelschen. Mannheim. Zarqba, A. (1965). Z zapomnianych wyraz6w polskich: (z)golemy. Prace filologiczne 18 (4): 199-206. Prispelo januarja 1996, sprejeto decembra 1996 Received January 1996, accepted December 1996 Zakaj je moino, da so obstajali stiki med Slovenci in iidi pred 1. 1000 n. 9. ObiEajno se domneva, da so se iidje do konca prvega tisotletja selili iz Sredo- zemlja prek severne Francije in Italije na Bavarsko in v Porenje. Judaiziran! nemika naretja (znana kot jidig) naj bi se oblikovala, ko so ti romansko govoreti Zidje pre- vzeli nemgka naretja. Tako se jidig obiEajno pojmuje kot judaizirana oblika visoko- nemgkih naretij. V prispevku jevpokazano, kako jezikoslovni dokazi omogotajo vzpostavitev nove teorije o selitvi Zidov (tako palestinskih emigrantov kot domorod- nih evropskih spreobrnjencev), namreE da so se do poznega prvega tisotletja n. 9. prek Balkana selili na dvojezitna luiigkosrbsko govoreEa ozemlja. S tega staligta bi morali jidig oblikovati balkanski iidovski emigranti, ki so govorili enega ali vet balkanskih jezikov: romansko, grgko inlali juinoslovansko. V nasprotju z obitajnim pojmovanjem je na jidig najbolje gledati kot na naretje luiigtine, ki je bilo naposled preleksikalizirano s srednjevisokonemgkim besedjem (kar je na koncu imelo za po- sledico slovanski jezik s prevladujotim nemgkim besedigtem). Prispevek poudarja dokaze, ki izhajajo s slovenskega ozemlja. 68 Slovenski jezik - Slovene Linguistic Studies 1 (1997) Why There May Have Been Contacts between Slovenes and Jews before 1000 A.D. The traditional view of Jewish settlement history in Europe posits migrations from the Mediterranean region through northern France and Italy into Bavaria and the Rhineland by the end of the first millennium. The Judaized dialects of German (known as Yiddish) were allegedly created when these Romance-speaking Jews switched to regional German dialects. Yiddish has traditionally been defined as a Judaized form of High German dialects. This paper will demonstrate how linguistic evidence allows us to postulate an innovative theory about the migration of Jews (both Palestinian emigres and indigenous European converts) - namely, through the Balkans into the bilingual Sorbo-German lands by the late first millennium A.D. In this view, Yiddish was created by Balkan Jewish emigres speaking one or more Balkan languages: Romance, Greek and/or South Slavic. Contrary to the common view, Yiddish is best seen as a dialect of Sorbian which eventually became relexified to Middle High German (thus resulting in a Slavic language with an overwhelming German vocabulary). The article specifically emphasizes evidence from the Slovene territory.